Class V2vModel

A voice conversion model. This is a model that can be used to convert audio from one voice to another.

Implements

Constructors

constructor

Properties

client createdAt creatorDisplayName creatorGravatarHash creatorSetVisibility creatorUserToken creatorUsername fetchProfile ietfLanguageTag ietfPrimaryLanguageSubtag isFrontPageFeatured modelType title token updatedAt username

Methods

#fetchInference #getAudioUrl #uploadAudio infer

Constructors

constructor

  • new V2vModel(client, data): V2vModel

  • Parameters

    • client: Client

      The main client.

    • data: {
          created_at: Date;
          creator: {
              display_name: string;
              gravatar_hash: string;
              user_token: string;
              username: string;
          };
          creator_set_visibility: "public" | "hidden";
          ietf_language_tag: string;
          ietf_primary_language_subtag: string;
          is_front_page_featured: boolean;
          model_type: string;
          title: string;
          token: string;
          updated_at: Date;
      }

      The raw model data from the FakeYou API.

      • created_at: Date
      • creator: {
            display_name: string;
            gravatar_hash: string;
            user_token: string;
            username: string;
        }
        • display_name: string
        • gravatar_hash: string
        • user_token: string
        • username: string
      • creator_set_visibility: "public" | "hidden"
      • ietf_language_tag: string
      • ietf_primary_language_subtag: string
      • is_front_page_featured: boolean
      • model_type: string
      • title: string
      • token: string
      • updated_at: Date

    Returns V2vModel

Properties

Readonly client

client: Client

Readonly createdAt

createdAt: Date

Readonly creatorDisplayName

creatorDisplayName: string

Readonly creatorGravatarHash

creatorGravatarHash: string

Readonly creatorSetVisibility

creatorSetVisibility: "public" | "hidden"

Readonly creatorUserToken

creatorUserToken: string

Readonly creatorUsername

creatorUsername: string

fetchProfile

fetchProfile: ((this) => Promise<ProfileUser | undefined>) = implFetchUser

Fetch the profile of the user who created this model.

Type declaration

    • (this): Promise<ProfileUser | undefined>

    • Fetch the profile of the user who created this model.

      Parameters

      Returns Promise<ProfileUser | undefined>

Readonly ietfLanguageTag

ietfLanguageTag: string

Readonly ietfPrimaryLanguageSubtag

ietfPrimaryLanguageSubtag: string

Readonly isFrontPageFeatured

isFrontPageFeatured: boolean

Readonly modelType

modelType: string

Readonly title

title: string

Readonly token

token: string

Readonly updatedAt

updatedAt: Date

Readonly username

username: string

Alias

creatorUsername

Methods

Private #fetchInference

  • #fetchInference(uploadToken): Promise<undefined | {
        inference_job_token: string;
        success: true;
    }>

  • Parameters

    • uploadToken: string

    Returns Promise<undefined | {
        inference_job_token: string;
        success: true;
    }>

Private #getAudioUrl

  • #getAudioUrl(inferenceJobToken): Promise<undefined | {
        created_at: Date;
        job_token: string;
        maybe_result: {
            entity_token: string;
            entity_type: string;
            maybe_public_bucket_media_path: string;
            maybe_successfully_completed_at: null | Date;
        };
        request: {
            inference_category: string;
            maybe_model_title: null | string;
            maybe_model_token: null | string;
            maybe_model_type: null | string;
            maybe_raw_inference_text: null | string;
        };
        status: {
            attempt_count: number;
            maybe_assigned_cluster: null | string;
            maybe_assigned_worker: null | string;
            maybe_extra_status_description: null | string;
            maybe_first_started_at: null | Date;
            requires_keepalive: boolean;
            status: "complete_success";
        };
        updated_at: Date;
    }>

  • Parameters

    • inferenceJobToken: string

    Returns Promise<undefined | {
        created_at: Date;
        job_token: string;
        maybe_result: {
            entity_token: string;
            entity_type: string;
            maybe_public_bucket_media_path: string;
            maybe_successfully_completed_at: null | Date;
        };
        request: {
            inference_category: string;
            maybe_model_title: null | string;
            maybe_model_token: null | string;
            maybe_model_type: null | string;
            maybe_raw_inference_text: null | string;
        };
        status: {
            attempt_count: number;
            maybe_assigned_cluster: null | string;
            maybe_assigned_worker: null | string;
            maybe_extra_status_description: null | string;
            maybe_first_started_at: null | Date;
            requires_keepalive: boolean;
            status: "complete_success";
        };
        updated_at: Date;
    }>

Private #uploadAudio

  • #uploadAudio(file): Promise<undefined | {
        success: boolean;
        upload_token: string;
    }>

  • Parameters

    • file: Buffer

    Returns Promise<undefined | {
        success: boolean;
        upload_token: string;
    }>

infer

  • infer(audio): Promise<undefined | V2vAudioFile>

  • Infer uploaded audio to a new voice represented by this model.

    NOTE: Voice conversion models are very slow to infer. It may take multiple minutes for the inference to complete.

    Parameters

    • audio: Buffer

      The audio to infer as a wav file.

    Returns Promise<undefined | V2vAudioFile>

    The inferred audio file. Undefined if the audio could not be inferred.

Settings

Member Visibility

Theme

On This Page

Generated using TypeDoc